Neviong

The Neviong (Pronounced "Nev-e-ong") Internet-Entertainment-Television-Game A convergence device with a gaming capability using emulation to give access to thousands of titles.

This is an informal proposal I put to a company a while back (it was also posted on-line at their request). They weren't in a position to do anything about it at that time but since it was made public I'm posting this edited version here.

The general idea for this has been around for quite some time but this expands on how to go about the project making suggestions for the OS applications and hardware. It is not written in a formal manner, more of a "brain dump".

Notes:

1) The device is intended for sitting under the TV rather than acting as a computer though it could be used for both.

2) It's also based on the idea of a dedicated board developed for the purpose (not described here) but the additional functions could be added to a standard PC via add in cards though there would be some interesting problems to solve if that was done.

3) I had some ideas for the GUI but I'm doing this as a private project.

Proposal for "The Neviong"

A consumer electronics device with STB, media recording / playback, internet and gaming capabilities. Intended for licensing to 3rd party manufacturers along with access to an on-line game collection.

This system is a combination of both hardware and software, by providing both it can serve as a turn-key solution for manufacturing companies who do not have the R & D departments to develop such a machine. Can also be modified to fit with a manufacturer's requirements - i.e. add their logo / colours to the GUI.

Despite the complexity of the system described below I don't think this will take a big team or take an excessive amount of time.

The initial specification / research / design and prototyping phase (software) can be done by one person. At a guess the initial phases will take a few months with other people coming in later. The system will be made from a number of primarily existing systems and the work will largely consist of modifying these to work with the control system and GUI of the system.

Software
I'd use Linux as the OS as this means much of the software necessary is already available as well as some high grade games (including Doom 3).

In order to produce a device with a large number of media and computing functions for the consumer, it is going to be a design challenge to say the least. For lot of the parts we can use existing open source applications but they need to be made to work in a seamless manner, with a single, simple GUI. All of this has to be done through a remote control, not keyboard or mouse.

This will require a pretty sophisticated control system which can interpret input then controls the hardware, applications and produces the correct display.

The current idea is to build the control system in Python on Linux using a web browser and standard web technologies (XHTML / JavaScript / CSS) to build the GUI. By using these technologies the system can be made very flexible and portable and can thus be used for different devices and products. Providing Python is supported the system could run on MorphOS though the type of HTML used will be limited with current browsers.

As this will be an international product the displays will need to be created in the correct language, the user should also be able to change this.

As this is to be a product for licensing different manufacturers will want to customise the GUI so this should also be possible.

3rd Party Licensed Software Games will of course be 3rd party but we will also need access to system ROMs. The system should be capable of running (through emulation) all pre-PC computer games and console games up to PS1 / N64, The GameCube may also be a possibility if a PowerPC CPU is used.

MorphOS *might* be usable for Amiga emulation with work - if not we could see about a special version of Amiga Forever or even go direct and talk to KMOS.

These can be sold as CDs or via an on-line service.

Other external software will also need to be licensed, in some cases this is existing open source software but licenses will still be needed: MHP, Java, media codecs, DVD etc.

While the system is based on existing software the aim is to get it to work together in a seamless manner as would be expected from a consumer electronics device. The user should not know that this is really a computer.

The Control System
This is the basis of the system, it is the glue that binds all the parts together and generates the GUI displays. This is the part which will be difficult to design but this in itself means it's a marketable product.

The control system combined with pre-licensed 3rd party packages and a hardware reference design gives manufactures a full solution for making their own neviong.

GUI

XHTML based.

Generated on the fly from templates.

GUI Requirements

Easy to use / Navigate

Multilingual.

Customisable for different manufacturers.

Easy to control using a remote control.

Generates display for LCD display (should be useable without a screen for audio playback & radio).

Other notes
DVD / CD drive needs to recognise media and data discs - a computer CD drive will not play many protected discs as they contain errors, in this case the Neviong should act like a CD player and ignore the errors.

Development Phases (Software only)

1) Requirements gathering

Complete list of desired features
Fully list all functionality required to perform these functions
Write a specification and requirements list

2) Design

Research (which components to use, licenses required etc.)
Architectural design
Initial interaction design

Design does not stop here, it continues to track changes and adapt in phases 2 and 3.

3) Construction 1

Full interaction and GUI design
Initial setup of control system (browser, server, page generation, input control)

4) Construction 2

Control system scripts
Application level changes
Integration

5) Beta Testing

As a consumer product this will have to be extensively tested.
Normal testing should be a continuous process - set a few machines aside and run a set of scripts on each new build.

6) Maintenance

This cannot be maintained like a normal system but options such as on-line updates and updater CDs should be considered.
It should be considered however that some machines will almost certainly never be updated - hence the aggressive testing and security control should be built in.

Hardware
The hardware will need to be a board custom built for this purpose*, for STB (Set-Top-Box) / convergence parts a standard STB processor will do but for gaming a much more powerful CPU and graphics system is required, Emulation also requires a lot of processing power and again a STB device will not be capable of this.

I can look into and make recommendations for parts we could use for the hardware. My current inclination would be to have two independent systems in one, one handles games, the other handles STB tasks. This has the side benefit of solving the problem of a bad program crashing the system if it is recording, with independent systems such a crash will be highly improbable.

For running emulations and modern(ish) games a CPU in the class of a 1GHz+ G4 or low end 970 will be required. A modern-ish GPU will be required with complete 3D drivers as well as an OpenGL subsystem. For the majority of the STB functionality a standard STB device can be used, this gives additional stability as critical parts can be completely independent from the gaming parts.

I think there's some partnerships possible for the hardware as the companies who produce the components may also be interested in using it as a reference design.

*For volume production a single board version would be necessary to get costs down, An existing Pegasos (PowerPC computer) with add-ons (i.e. TV card) could serve as an initial development system and possibly product.

Notes:

I made these comments are in answer to posts made following the above being put on-line.

1) The reason it's based on Linux is because a lot of the work is already done and out there to be used. I don't necessarily think it's so much of a big job as a complex job. The integration is the most difficult part, that will require a development team, but shouldn't take very long.

2) This would have STB functionality but I wouldn't call it a STB. It's more of one of those mythical "convergence" boxes combined with a Games Console. Way more than any STB. It's competition is Microsoft's media centre PCs.

It'll be expensive (at least initially) but todayÕs high end parts are tomorrowÕs low end.

It needs a fairly high end G4 for the games as emulators take a lot of power, otherwise you end up with C64 style games and thatÕs it, by including a decent processor and Graphics chip it can emulate a multitude of different systems and thus from day one has *thousands* of games available, some of which will be very well known.

It also needs to be made in volume which is why I like the idea of doing it as a reference design, one company develops it and someone with manufacturing expertise - and the money - makes it.

I liked the idea of the Neviong as everyone is making a convergence box but I doubt anyone other than Sony and Microsoft is going to make one with thousands of games so this is a way for other manufactures to compete, of course this can all hook up with an online system for downloading games - want to be the Microsoft of convergence?.

Adding thing like webcams etc. adds functionality so makes it potentially even more useful.

Back to The Source.
Nicholas Blachford's home page.